Hello all,

 

Every now then and then i get this hardware error in syslog. I don't know how to use "edac_mce_amd module instead"

 

Mar 21 07:39:17 LoweNas kernel: mce: [Hardware Error]: Machine check events logged Mar 21 07:39:17 LoweNas kernel: [Hardware Error]: Deferred error, no action required. Mar 21 07:39:17 LoweNas kernel: [Hardware Error]: CPU:1 (19:21:2) MC18_STATUS[-|-|-|-|-|-|Deferred|-|-]: 0x9090909090909090 Mar 21 07:39:17 LoweNas kernel: [Hardware Error]: IPID: 0x0000000000000000 Mar 21 07:39:17 LoweNas kernel: [Hardware Error]: Bank 18 is reserved. Mar 21 07:39:17 LoweNas kernel: [Hardware Error]: cache level: RESV, tx: INSN Mar 21 08:33:51 LoweNas root: Fix Common Problems: Error: Machine Check Events detected on your server Mar 21 08:33:51 LoweNas root: mcelog: ERROR: AMD Processor family 25: mcelog does not support this processor. Please use the edac_mce_amd module instead.

lowenas-diagnostics-20250321-0843.zip

you are running a AMD machine and most likely your ram stick failed in a task but ECC Ram fixed it. so something is loged.

This could be early signs of the CPU memory controller failure or physical ram hardware issues.

Please reboot unraid and at the grub screen run mem test.

 

this seems like early signs of ram failure.

It completed the ram test no errors. is this somthing i need to add somewhere "edac_mce_amd module instead"

 

thanks for the help

unfortuantely no the edac_mce_amd module is due to your motherboard and the software. mcelog has been broken on amd platform for a number of releases...

 

On AMD systems, the mcelog service is often superseded by the edac_mce_amd kernel module for handling hardware error logging, and the mcelog service may report errors if it's not designed for the AMD processor family. 
Here's a more detailed explanation:
mcelog and AMD:
The mcelog daemon is designed to log and handle hardware error messaging on x86_64 platforms, but on AMD systems, the edac_mce_amd kernel module takes over this function. 
edac_mce_amd:
This kernel module is responsible for handling machine check exceptions (MCes) on AMD processors, logging errors into the system logs instead of mcelog's log file. 
Error Messages:
If you encounter an error message like "mcelog: ERROR: AMD Processor family XX: mcelog does not support this processor. Please use the edac_mce_amd module instead," it indicates that mcelog is not the correct tool for logging hardware errors on your AMD system

If memetest came back fine, you can then safely ignore the error. as 90% of the time this is on AMD and how the amd reset bug / memory controller triggered ecc memory operation and was flagged by the memory check event log. If it continues to happen, this is early signs of ram / cpu memory controller failure.

Limetech / Unraid is aware of the amd mce issue and error. the module is not installed in Slackware and would require a kernel level edit to add and use. Doing so would break it for intel as well. More of limetech business is on intel and apples server hardware... 

I have a feature request and other reports on it lost on the forum. i'm not the first, i won't be the last. if mem test failed, you now need to find what hardware component failed and why. replace it if failed.

 

 

 

 

https://access.redhat.com/solutions/158503

 

also mcelog is old / deprecated, you should use rasdaemon instead... but rasdaemon don't have a slack pack or build its debian only...